Uric Acid and Chronic Kidney Disease: Still More to Do

Abstract

Gout and hyperuricemia are present in 25% and 60% of patients with chronic kidney disease (CKD), respectively. Despite the common association, the role of uric acid in the progression of kidney disease and in metabolic complications remains contested. Some authorities argue that the treatment of asymptomatic hyperuricemia in CKD is not indicated, and some have even suggested hyperuricemia may be beneficial. Here, we review the various arguments both for and against treatment. The weight of the evidence suggests asymptomatic hyperuricemia is likely injurious, but it may primarily relate to subgroups, those who have systemic crystal deposits, those with frequent urinary crystalluria or kidney stones, and those with high intracellular uric acid levels. We recommend carefully designed clinical trials to test if lowering uric acid in hyperuricemic subjects with cardiometabolic complications is protective.

Hyperuricemia (defined as a serum uric acid level >7 mg/dl in males and >6 mg/dl in women) is common in CKD. This is because hyperuricemia is common in type 2 diabetes and hypertension, which are conditions that cause CKD, and also because CKD results in reduced urinary excretion of uric acid. As a consequence, the prevalence of gout increases from 1% to 2% of adults with normal kidney function to 32% of those with stage 4 CKD. Hyperuricemia prevalence also increases from 11% among those with normal kidney function to 80% among those with stage 4 CKD.¹ The converse is also true. CKD stage 2 or higher is present in 70% of subjects with gout and 50% of those with hyperuricemia.²

It is therefore important to understand if high uric acid levels modify kidney or metabolic outcomes. This is especially true because hyperuricemia is an independent predictor of CKD and metabolic diseases,3, 4, 5 including in subjects who are healthy without any morbidities.⁶ There is also a direct relationship of serum uric acid level with prevalence of hypertension, diabetes, and CKD (Figure 1).^7,8

Figure 1.

Relationship of Serum Uric acid with CKD. (a) A study in which more than 48,000 Japanese that were 20 years or older who were followed for 7 years. After controlling for baseline serum creatinine and other variables, the presence or absence of baseline hyperuricemia (defined as >7 mg/dl in men and >6 mg/dl in women) markedly increased the risk for developing end stage kidney disease requiring dialysis. (b) A figure based on the study of 5707 participants aged 20 years and older from the National Health and Nutrition Examination Survey 2007–2008. There is an exponential relationship of serum uric acid levels with CKD. (a) Adapted from Iseki et al.⁷ and (b) Adapted from Zhu et al.⁸ CKD, chronic kidney disease; HyperUric, hyperuricemia.

Today, there remains controversy over the role of uric acid in CKD and cardiometabolic outcomes. Several groups have suggested that asymptomatic hyperuricemia in CKD is benign and should not be treated, or may even be beneficial.9, 10, 11, 12 Here, we present our countering viewpoint with recommendations for how to move forward. For purposes of the discussion, our analysis will be separately about those who have gout and CKD, and those with asymptomatic hyperuricemia and CKD.

Gout and CKD

Gout is classically treated with urate-lowering agents to reduce the risk for recurrent arthritic attacks and joint damage.¹³ It has remained controversial whether lowering serum uric acid in gout has an effect on kidney disease or cardiovascular events. However, there are at least 3 arguments that strongly suggest that lowering uric acid in subjects with gout may be beneficial for CKD and/or cardiovascular events. All these are based on the fact that urate crystals are known to be very proinflammatory, and known to induce local inflammation that is mediated by activation of inflammasomes and the release of interleukin-1.¹⁴

The first argument is that urate crystals are known to deposit not only in joints, but also in other tissues,¹⁵ and one of the favored sites is in the collecting ducts of the kidney.¹⁶ As urine concentrates, it also acidifies, and this can lead to urate crystallization. Some urate crystals adhere to the tubular epithelium, where they can cause local inflammation that leads to rupture of the tubular wall with the crystals escaping into the interstitium.¹⁷ This can be associated with marked local inflammation with macrophage infiltration.¹⁸ Subjects with gout have reduced fractional excretion of uric acid. Nevertheless, they excrete large amounts of uric acid, especially following ingestion of a purine-rich¹⁹ or fructose-rich²⁰ meal.

Indeed, in the days before urate-lowering treatment was available, autopsies of patients with gout showed an almost universal presence of kidney disease, characterized by arteriolosclerosis, focal segmental glomerulosclerosis, and chronic tubulointerstitial disease,^16,21 and as many as 20% or 25% of gouty subjects would have markedly reduced kidney function.²² Most strikingly, urate crystals were found in 90% of autopsies, always concentrated in the outer medulla.²¹ In contrast, kidney biopsies rarely document urate crystals²³ because the biopsies are of the cortex where uric acid crystals are sparse. In addition, urate crystals are often washed out during the fixation process unless special techniques, such as using alcohol fixation and the De Galantha stain, are used.

It is not known how common urate crystals are present in the kidneys of gout patients today. Newer techniques such as the dual energy computed tomography (DECT) scans have been suggested as a method to detect their presence.²⁴ However, there are technical issues in performing and analyzing scans, and adjustments need to be performed for each organ evaluated, and the art of performing DECT scans on kidneys is still being refined. Nevertheless, another approach that has been recommended is the use of renal artery ultrasound to determine if there is enhanced echogenicity in the renal medulla associated with crystal deposition. A hyperechoic medulla by ultrasound is considered strongly suggestive of urate crystal deposition and has been reported to be present in one-third of subjects with gout.²⁵

The second argument to treat gout not just for the risk of recurrent gout attacks is because there is now evidence that urate crystals may be directly involved in the atherosclerotic process. Specifically, DECT scans adjusted for evaluating blood vessels have had the surprising discovery that urate crystals are common in the aorta and coronary arteries of subjects with gout. Indeed, urate crystals are present in the blood vessels of 75% to 86% of patients with gout, with nearly 30% of gout subjects having crystals in their coronary arteries.^26,27 The primary sites seem to be in areas of atherosclerotic plaque, which has been confirmed by histologic studies.28, 29, 30 Urate crystals have also been colocalized with sites of vascular calcification.²⁶ Urate crystals are likely to stimulate inflammasomes in the lesions similar to that of cholesterol crystals, and this is expected to increase the risk for plaque extension or rupture.³¹ These findings could explain why both hyperuricemia and gout are associated with cardiovascular mortality in both epidemiology and Mendelian randomization studies.32, 33, 34

The third argument relates to the observation that urate crystal deposits can result in not just local inflammation but also in systemic inflammation, and the latter is recognized as a contributing factor for progressive kidney disease as well as cardiovascular events.^35,36 A striking finding in gout is that resolution of an acute attack is not associated with resolution of urate crystals because they will persist until the next attack (i.e., the “intercritical period”).³⁷ These “silent crystals” can still be associated with evidence for systemic inflammation that may have an indirect role in the progression of kidney disease and cardiovascular events.³⁸ Indeed, both elevated monocyte counts and high levels of highly sensitive C-reactive protein levels are found in patients with a history of gout.^39,40 Some studies have found that allopurinol treatment can lower highly sensitive C-reactive protein levels.⁴¹ Furthermore, targeting inflammation by giving antibodies to interleukin-1 can reduce highly sensitive C-reactive protein and cardiovascular events in individuals at risk of cardiovascular disease.^42,43

We believe these arguments are strong enough to recommend urate-lowering in all subjects who suffer from gout, and treatment could be dietary or involve a nutraceutical or drug; however, the goal would be to lower serum uric acid levels to <6 mg/dl.

Hyperuricemia and CKD

Initially there was strong evidence that hyperuricemia in the absence of gout might be driving CKD. This was supported both by epidemiologic studies,^3,44 experimental studies^45,46 and pilot clinical trials.⁴⁷ However, 2 large clinical trials, known as the prevention of early renal loss in type 1 Diabetes and CKD-FIX (Controlled Trial of Slowing of Kidney Disease Progression from the Inhibition of Xanthine Oxidase), were published in the same issue of the New England Journal of Medicine and found no benefit of allopurinol in slowing renal progression.^48,49 This initially led several groups, including the Caring for Australians and New Zealanders with Kidney Impairment Guidelines Committee to suggest that there is now conclusive evidence not to treat asymptomatic hyperuricemia in CKD.^9,12

We believe this conclusion is premature. The prevention of early renal loss and CKD-FIX studies tested whether lowering uric acid was beneficial but not whether treating hyperuricemia is beneficial, because both studies included large numbers of patients with normal serum uric acid levels. Normal uric acid levels are not expected to significantly increase the risk for CKD (see Figure 1).

Both studies were also intention-to-treat analyses. These analyses count all treated subjects even if they discontinued treatment because of problems with compliance or side effects. In this case, approximately 17.5% to 30% of subjects dropped out before completion of the trial. The fact that this happened in both treatment and placebo groups suggest it was largely not because of the drug but rather because of general characteristics of the population being studied, or perhaps because of general concerns or preset opinions that allopurinol might be associated with high side effects. Therefore, the trial was not actually testing the hypothesis that asymptomatic hyperuricemia might be driving CKD, but rather tested whether allopurinol treatment reduced progression of kidney disease when including issues such as compliance, side effects, and other factors.

The CARI guidelines have some additional problems. For example, the criteria they used in their studies did not require the presence of hyperuricemia (allowing any serum uric acid level), nor did it require the presence of CKD,⁹ yet the question being addressed was whether treating asymptomatic hyperuricemia in CKD was beneficial. Their conclusion that treatment was not beneficial was also at odds with the meta-analysis they used for their study. The latter had concluded that lowering serum uric acid was beneficial in reducing the decline in estimated glomerular filtration rate and blood pressure.⁵⁰

More recently, the ALLHEART study was published in which older patients (age >60 years) with a history of ischemic heart disease and no gout were randomized to allopurinol or placebo and followed for almost 5 years without any discernible benefit on subsequent cardiovascular events.⁵¹ However, similar to the other trials, patients with normal uric acid levels were included (the mean serum uric acid was 5.6 mg/dl) and gout was excluded. Likewise, there was a large (57%) dropout which were included in the analysis because it was an intention-to-treat study, such that it was not truly testing whether the treatment of hyperuricemia is beneficial on cardiac endpoints.

One might hope that meta-analyses might help resolve the issues, but even here there remains confusion, because some recent studies suggest that treatment of asymptomatic hyperuricemia in CKD does slow progression,^52,53 whereas others are mixed or indeterminant,^50,54,55 and some are negative.⁵⁶ Some meta-analyses show benefit with only specific urate-lowering drugs, such as febuxostat.^57,58 One possible explanation for the mixed data is that there may be subgroups that particularly benefit from treatment. In the next section, we propose specific groups that we believe might be most likely to respond to urate-lowering therapy.

Subgroups in Which Asymptomatic Hyperuricemia may be Most Likely to Drive CKD and Cardiac Disease

The first group to consider would be patients with asymptomatic hyperuricemia who may harbor “silent” crystals in their joints, blood vessels, or kidneys (similar to the argument provided in the gout section). Approximately 15% of hyperuricemic subjects who do not have gout still have urate deposits in their blood vessels when evaluated by DECT scan.²⁶ Some patients may also harbor crystals silently in their joints or kidneys. Studies that include DECT scanning or renal ultrasound may help identify this subgroup.

A second group would be subjects with recurrent urate crystalluria or kidney stones. Crystalluria can stimulate inflammasomes in renal tubular cells leading to local inflammation and injury that can accelerate CKD.^11,59, 60, 61 One mechanism for the crystalluria would be the presence of a low urinary pH, such as may occur with dehydration, because uric acid is very insoluble in acidic urine. A low urinary pH predicts the development of CKD.⁶² This may also explain the benefit of bicarbonate therapy to slow CKD, and we reported that bicarbonate therapy can solubilize urate crystals and reduce biomarkers of renal tubular injury in diabetic subjects.⁶³ Heat stress-associated urate crystalluria is also common in subjects with Mesoamerican nephropathy where it might be driving kidney damage,^64,65 and uricosuria accompanies rhabdomyolysis where it has been suggested to play an ancillary role.⁶⁶ Indeed, both allopurinol and bicarbonate therapy are protective in experimental rhabdomyolysis-associated kidney injury.^67,68

Although urate crystalluria is largely driven by low urinary pH, some individuals will show “overproduction” uricosuria. Recently, this was shown to be mediated by reduced intestinal uric acid excretion, especially by inhibition of the adenosine triphosphate-binding cassette subfamily G member 2 urate transporter.^69,70 The adenosine triphosphate-binding cassette subfamily G member 2 transporter is expressed in both the kidney and intestine, but polymorphisms that are associated with reduced function (such as the Q126X variant [that has near absent function] and the Q141K [which has half-function]) result in reduced intestinal elimination with increased renal excretion.⁷⁰ New studies suggest that individuals carrying these variants are at higher risk for progression of CKD.^69,71 Dietary mechanisms may also be operative. Fructose, for example, also blocks adenosine triphosphate-binding cassette subfamily G member 2 in the intestine, reducing intestinal uric acid excretion,⁷² while increasing urinary excretion,⁷³ reducing urinary pH,⁷⁴ and reducing urine volume.⁷⁵ Dietary intake of high purine foods can also cause transient uricosuria¹⁹ that might be important in mediating kidney injury.

Asymptomatic hyperuricemic patients with kidney stones may also be candidates for having recurrent uricosuria. Kidney stones are common in subjects with gout, being present in one-third of subjects when evaluated by helical computerized tomography, although two-thirds of these patients were not aware they had stones. Of interest, the subjects with kidney stones had worse kidney function and lower urine pH than those who did not have stones.⁷⁶ Therefore, one might consider evaluating subjects with hyperuricemia to determine if they have stones, because this may represent a group with an increased risk for CKD progression.

The importance of urate crystalluria in causing kidney injury was recently elucidated in patients with uricosuria and hypouricemia because of the loss of urate transporters in their kidneys.^77,78 In both experimental models and humans, the use of allopurinol to reduce uricosuria is associated with protection from kidney injury.^77,79

A third group of subjects might be those who have increased intracellular levels of uric acid occurring in their livers or kidneys. In the liver, intracellular uric acid is thought to mediate oxidative stress to the mitochondria that drives metabolic effects like insulin resistance, hepatic fat accumulation, and elevation in blood pressure.⁸⁰ Interestingly, this is associated with dietary measures such as intake of fructose or purine-rich foods; howver, it can also occur with block in intestinal uric acid excretion.⁷¹ Another condition might by polycystic kidney disease, in which the enlarging cysts are likely stimulating local uric acid generation. In autosomal polycystic kidney disease, serum uric acid levels are high and correlate with progression, whereas pilot studies suggest that lowering serum uric acid may be protective.^81,82

A major mechanism by which hyperuricemia induces its effects on the kidney is likely via its vascular effects. Uric acid has been shown experimentally to mediate endothelial dysfunction by both reducing biologically available nitric oxide and inducing oxidative stress.83, 84, 85 Although not all studies are positive, most clinical studies suggest that xanthine oxidase inhibitors can improve endothelial dysfunction.⁸⁶ Uric acid can also enter into vascular smooth muscle cells via specific transporters,^87,88 where it drives proliferation and proinflammatory pathways.^87,89,90 Hyperuricemia is especially correlated with disease of the afferent arteriole in both experimental animals⁴⁵ and in humans.⁹¹ Experimentally, this is associated with renal vasoconstriction, altered renal autoregulation, and increased glomerular hydrostatic pressure.⁹² Serum uric acid also correlates with high renal afferent arteriolar resistance in humans.⁹³

Measuring intracellular uric acid levels is difficult. However, determining plasma xanthine oxidase activity might be an alternative way to identify these patients. Elevated plasma xanthine oxidase activity has been reported to identify subjects with CKD who are at risk for cardiovascular events⁹⁴ and may be superior to serum uric acid in predicting metabolic disorders.^95,96

Other Arguments

Some have argued that uric acid may be beneficial because it can function as an antioxidant.⁹⁷ However, uric acid is pro-oxidative inside cells because it activates nicotinamide adenine dinucleatide phosphate oxidase and also because it can generate radicals such as hydroperoxide, peroxynitrite-radicals, and myeloperoxidase-based radicals.^98,99 When uric acid quells the production of hypochlorous acid by neutrophils, the overall oxidative stress remains the same because of increased production of superoxide.¹⁰⁰ Indeed, the overall effects of soluble uric acid are proinflammatory.^98,100, 101, 102

Some groups have reported that hyperuricemia induced by inosine may protect against kidney disease.^11,103,104 However, inosine has been shown to be anti-inflammatory because it activates adenosine receptors105, 106, 107 and generates hypoxanthine, the latter that can be recycled to IMP and eventually adenosine triphosphate.¹⁰⁸ Importantly, in the studies evaluating if uric acid is beneficial in kidney disease, the investigators did not determine if allopurinol treatment would reverse this protection. Other groups have shown that the anti-inflammatory effects of inosine are enhanced by administering allopurinol.^109,110

Another issue has been the concern that febuxostat might increase the risk for cardiovascular mortality because it was associated with more cardiovascular events in the CARES trial.¹¹¹ Other studies could not affirm this association.¹¹² Moreover, there was no placebo group in the CARES study, and studies of allopurinol suggest it may reduce mortality risk in the general population,¹¹³ with a trend toward protection in subjects with CKD.¹¹⁴

An additional argument is that most Mendelian randomization studies have not been able to show that genetic polymorphisms that increase serum uric acid translate into increased risk for CKD,115, 116, 117, 118 despite this being shown in other Mendelian randomization studies for hypertension,¹¹⁹ coronary artery disease,³⁴ or cardiovascular events.³³ However, serum uric acid is not the critical factor, given that the factors driving kidney disease may relate more to urine uric acid, urine pH, presence or absence of urate crystals in the kidney, and intracellular serum uric acid levels, especially in the liver and kidney. Although intracellular uric acid and serum uric acid levels are often correlated, they can also be dissociated.¹²⁰ In addition, some urate transporters, such as SLC2A9, may have divergent effects on serum uric acid depending on which target organ is evaluated,^121,122 suggesting that genetic polymorphisms altering SLC2A9 function could have opposing biologic effects depending on the target organ that could confound Mendelian randomization studies.

Finally, an important consideration is that what initiates disease might be different from what drives it (Figure 2). For example, experimental studies suggest that experimental hyperuricemia causes a rise in blood pressure because of effects of uric acid; however, over time there is the induction of an autoimmune inflammatory response in the kidney that maintains the hypertension.¹²³ Likewise, experimental hyperuricemia is associated with glomerular hypertension and renal vasoconstriction that is dependent on uric acid levels, but as CKD develops, glomerular hypertension is driven by a reduction in nephron number.¹²⁴ Similarly, there is evidence that fructose-induced hyperuricemia initially causes a reversible insulin resistance; however, overtime there is progressive injury to the islets resulting in islet dysfunction and a loss in insulin secretory ability.¹²⁵ There is also evidence that longstanding mitochondrial oxidative stress may lead to a loss of mitochondria that may have persistent effects.¹²⁶ Thus, the timing for when uric acid is lowered may be important. Similarly, to see a benefit in established disease, one might have to treat for a long time, and there is some evidence that the benefit of lowering uric acid is greatest if treatment is prolonged for 2 years or more.¹²⁷

Figure 2.

Uric acid may be more Important in the Initiation of Metabolic Diseases Rather than the Maintenance. AMPK, adenosine monophosphatase-activated protein kinas; ATP, adenosine trisphosphate; CKD, chronic kidney disease; NO, nitric oxide; RAS, renin angiotensin system.

Summary

Asymptomatic hyperuricemia is common in subjects with CKD. Although the CKD-FIX and prevention of early renal loss studies did not report benefit of allopurinol in slowing the progression of kidney disease, they did not specifically address the role of hyperuricemia. We suggest that there may be some specific subgroups of subjects with asymptomatic hyperuricemia that would benefit, including those with documented crystal deposition in joints, blood vessels, and the kidneys; those with documented recurrent urate crystalluria or with kidney stones; and those who have evidence for elevated liver or kidney uric acid levels (possibly noted by high plasma xanthine oxidase activity). Examples of some proposed trials are shown in Figure 3. The addition of these carefully designed studies are needed to determine the role of uric acid in the progression of CKD.

Figure 3.

Examples of potential clinical trials to investigate the role of uric acid in cardio-renal diseases. BP, blood pressure; CRP, C-reactive protein; CV, cardiovascular; DECT, dual energy computed tomography; XO, xanthine oxidase.

Disclosure

RJJ has equity with Colorado Research Partners LLC, has stock with XORTX Therapeutics, and is a consultant for Horizon Pharma. MAL and LGl have equity with Colorado Research Partners LLC. CB has research grants from Servier and Alfasigma and serves as a consultant for Servier, Novartis, Novo Nordisk, Alfasigma, Sanofi, and Menarini. FP declares no competing interests.